58 BULLETIN 832, U. S. DEPARTMENT OF AGRICULTURE. 
eter of the channel, as may be seen by referring to Plate XX, figure 
2. The values of n obtained are very high, due no doubt principally 
to the presence of vegetation. 
Unusual care was exercised in making the slope measurements 
and gaugings for both courses of channel. The slope for course A 
was very small and required precise measurements to secure reli- 
able results. 
DISCUSSION OF FLORIDA EXPERIMENTS, 
Owing to the fact that the slopes for channels in this section of the 
country are quite small, the velocities are necessarily low. As a ~ 
result, the channels are subject to silting and to the rapid growth of 
vegetation. Unless channels in this section are to be carefully main- 
tained, it can be seen from the results of the experiments as given 
in Table 6 that a comparatively high value of n should be used in 
their design. 
VARIATION OF » WITH STAGE IN CHANNEL. 
The results of these experiments show that generally the value of 
nm decreases as the stage in the channel increases. This was pointed 
out in an article in the Engineermg News, February 24, 1916, by 
R. E. Horton, as being true for natural river channels and was 
attributed to the fact that: 
In determining the coefficients of roughness in natural river channels the entire 
cross section as determined by soundings is taken into account, whereas if the cross 
section is irregular there may be pockets or holes in the bed of the stream containing 
slack water at lower stages, so that the full area of section is not effective. Similarly at 
low stages of the stream there is likely to be more or less fall that is not effective as 
slope and that does not contribute to producing velocity. 
While the above is also generally true for dredged channels, yet 
these experiments show a great many exceptions, one of which is that 
of Mud Creek, where the values of n obtained were practically the 
same for all stages in the channel, owing to the fact that there was 
no appreciable amount of ineffective slope or cross-section and practi- 
cally no difference in the degree of roughness of the channel for 
different stages (see Plate ITI, fig. i and Table1). Another exception 
which shows the very reverse of the general statement above is that 
of Bogue Phalia, where n increased for an increase of stage. This 
was due to the excellent condition of the lower part of the channel] 
as compared with that of the upper part (see Plate V, fig. 1, and 
Table 2). 
APPLICATICN OF RESULTS. 
It is impossible to describe with absolute accuracy: the conditions 
existing in a channel, such as degree of regularity, amount and nature 
of obstructions and vegetation, and unifermity of cross section. 
